Data masking can be used for Anonymization or Pseudonymization to comply with GDPR, paper Redaction and Preprocessing for NLP models.